Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Prescription Medication Costs

Prescription Medication Costs BackgroundStudies in the past 25 years have suggested that physicians are not familiar with the costs of common prescription medications.ObjectivesTo determine physician familiarity with the cost of common prescription medications and to determine the value physicians place on knowing information regarding the cost of medications.DesignSurvey.SettingSeven community-based family medicine residency teaching clinics in Iowa.ParticipantsTwo hundred five practicing resident and faculty physicians.InterventionsFrom a series of $10 price intervals (range, $0.01-$80.00), physicians were asked to select the interval containing the cash price of the medication to an uninsured patient for 50 medications commonly prescribed in outpatient family medicine clinics. Physicians were also questioned about the value of medication cost information to their practice.Main Outcome MeasuresThe percentage of correct responses and the mean pricing scores were calculated for each respondent and for all medications.ResultsOne hundred seventy-eight physicians responded (86.8%). Only 22.9% of the responses correctly identified the cost of the medication. More than two thirds (68.3%) of the responses underestimated the correct price interval. Branded drugs were underestimated in 89.9% of responses, while generic drugs were overestimated in 90.2% of responses. Overall, 64.4% of physicians believed they did not receive sufficient information in their practices regarding prescription drug costs, and nearly all (93.6%) reported that regular information on prescription medication costs would help them prescribe more cost-effectively.ConclusionsPhysicians are unfamiliar with the costs of medications they commonly prescribe, and they report that regular access to information on prescription medication costs would help them prescribe more cost-effectively.INCREASED SPENDING and use of prescription medications contribute to escalating costs of health care. Attempts to control drug costs, through implementation of formularies and mandatory generic substitution, have had little impact on reducing the economic burden of prescription drugs. In the next 10 years, spending on prescription drugs is anticipated to increase an average of 9.8%, and account for nearly 8% of all national health expenditures.With 60% of physician office visits resulting in a prescription being written,selection of cost-effective pharmacotherapy is critical to curb increasing drug costs.On average, Americans use about 10 prescriptions per year.Access to affordable medications remains limited for some patients. Medicare does not have a prescription drug benefit, and approximately 30% of prescriptions filled are purchased by uninsured patients.In one survey,85% of physicians reported that inability to afford medications was problematic for their patients. Drug costs clearly influence patient compliance with therapy, and the inability to pay for medications can adversely affect treatment outcomes.While prescription drug insurance may reduce costs for individual patients, the inappropriate selection of more expensive medications may actually increase overall costs to society.Patients with prescription drug insurance are known to use brand name drugs at a higher rate than those who pay cash, which may reflect, in part, the fact that prescription coverage helps to insulate these consumers from the additional costs of more expensive brand name drugs.The ability of the physician to select cost-effective therapy is, therefore, important not only for the patient but also for consumers and employers paying the cost of the health insurance as well.Prescription drug purchases are economically unique because they are a "directed" demand.Physicians direct the purchase through drug selection and determination of appropriateness, with minimal input from the consumer. As the primary target of prescription drug marketing, physicians influence the market by acting as "gatekeepers" who direct prescription drug use decisions. To select cost-effective therapies, physicians must have an adequate knowledge of prescription drug costs.Studiesover the past 25 years have documented that physicians are not familiar with prescription drug costs. These studies have limited generalizability, as the method by which the drugs were selected does not ensure physicians surveyed routinely prescribe those medications. The most common office visit is to a family physician, and family physicians write more than 30% of prescriptions annually in the United States.Family physicians are thus in a position to direct most prescription drug purchases, and should have a basic familiarity with the costs of medications they routinely prescribe. This study identifies commonly prescribed medications in family physician clinics and determines the familiarity of physicians with the costs of these medications and their views on the importance of medication cost information to their prescribing behavior.PARTICIPANTS AND METHODSSTUDY PARTICIPANTSPracticing faculty and resident physicians from 7 family medicine residency training programs geographically distributed throughout Iowa participated. All programs are community based, in towns ranging from 30 000 to 190 000 people, averaging 15 000 to 40 000 patient visits per year; there is wide diversity of patients from each site. The approximate insurance status is as follows: Medicaid, 15% (range, 10%-19%); Medicare, 15% (range, 7%-25%); managed care, 23% (range, 10%-30%); and uninsured, 25% (range, 6%-56%). Each residency program has an average of 20 to 30 residents, 8 to 12 faculty physicians, and 1 clinical pharmacist. Residents at each site spend time in the ambulatory setting according to guidelines set by the American Council of Graduate Medical Education Residency Review Committee for Family Practice. First-year residents spend at least half a day per week in the ambulatory outpatient clinic, while second- and third-year residents spend at least 2 and 3 half days per week in the outpatient clinic, respectively.STUDY PROCEDURESDuring January 1998, a daily audit of written prescriptions was performed by clinical pharmacist faculty members (M.E.E., M.W.K., J.D.H., J.M.S., L.M.B., C.D.L., J.K.F., and H.A.K.) at each site. A representative sample of 50 commonly prescribed drugs among all clinics was selected for the survey, from 423 different medications and 3635 total prescriptions audited. For each medication on the survey, an attempt was made to insert into the survey at least one comparator medication from the same therapeutic class that was also in the top 50 commonly prescribed agents from the audit. The final 50 drugs chosen for the survey represented 50.3% of all prescriptions audited during the study period.Medications were presented in random order on the survey to prevent any cost patterns from emerging, and quantities were chosen to approximate a 1-month supply for long-term medications or the standard course of therapy for short-term medications. Generic names were used, with brand names supplied for medications available only as single source, branded drugs. Physicians were instructed to assume the prescription would be dispensed as written, and then asked to select what they believed to be the correct prescription cost range for an uninsured patient from a series of $10 price intervals (range, $0.01-$80). Physicians were also asked to respond on a Likert scale to 4 statements regarding the importance of medication cost information to their practice. Statements included asking whether physicians believed they received adequate information regarding the costs of the medications they prescribe, whether regularly accessible information on medication costs would help improve their practice and help them prescribe more cost-effectively, and whether the prescription drug insurance status of patients influences their choice of medication to prescribe. The survey was distributed to faculty and residents during March 1998, which allowed first-year residents to gain almost 8 months of prescribing experience in the outpatient setting. The survey took approximately 15 minutes to complete.Because medication prices can vary greatly by type of pharmacy (chain or independent) and location, both of which could affect the scoring of the survey, the actual price for each medication was calculated using a standard formula used by Iowa Medicaid for reimbursement: branded drugs = [average wholesale price (published in the 1998 Red Book) − 10%] + $6.25, and generic drugs = maximum allowable cost set by the Health Care Financing Administration for multisource drugs + $4.02. Two local pharmacies were surveyed to compare the calculated price vs the price charged by the pharmacy. Discrepancies of more than $10 were resolved by using the pharmacy-reported price. This occurred in less than 6 cases. Correct price interval selections were coded with a value of zero. Incorrect selections were coded as the absolute value of the number of intervals above or below the correct interval.STATISTICAL ANALYSISAll statistical analyses were performed using statistical software (Statistical Product and Service Solutions, version 7.5.1; SPSS Inc, Chicago, Ill). Drugs were categorized as branded or generic, and as high (≥$60.01), medium ($20.01-$60.00), and low ($0.01-$20.00) cost for comparisons.The total number and the percentage of correct responses, as well as mean scores, were calculated for each respondent and for all medications. Mean scores were calculated to allow for more accurate reporting of the magnitude of the difference between price estimated by the physician and correct price interval. Because absolute values were used to code the number of price intervals above or below the correct interval, mean scores closer to zero indicate greater familiarity with price. Paired ttests were performed to identify differences in mean scores, and in percentage of correct responses for branded and generic categories. An analysis of variance with post-hoc Tukey correction for multiple comparisons was used to identify differences in mean scores and percentage of correct responses for each category, and for comparisons between faculty and resident classes. P<.05 was considered statistically significant.RESULTSThe survey was completed by 178 (86.8%) of 205 eligible physicians. Nonresponders to the survey generally were physicians who were on an off-site rotation at the time the survey was administered. Respondents were nearly equally distributed between faculty (n = 43) and first-year (n = 44), second-year (n = 46), and third-year (n = 45) residents. For faculty respondents, the mean (±SD) number of years since graduation from medical school was 17.4 (±7.9) years, with a mean (±SD) of 8.6 (±6.5) years as faculty.Overall, 68.3% of the responses underestimated the correct price interval for all drugs. Only 22.9% of the responses correctly identified the cost. The prices of branded drugs were underestimated in 89.9% of responses, while generic drugs were overestimated in 90.2% of responses. The medications whose prices were most commonly underestimated included fluoxetine hydrochloride (Prozac) (93.8% of responses underestimated), cefuroxime axetil (Ceftin) (82.6%), ciprofloxacin hydrochloride (Cipro) (87.0%), loratadine (Claritin) (87.0%), and omeprazole (Prilosec) (76.8%), while the medications most commonly overestimated included trazodone hydrochloride (94.4% of responses overestimated), a combination of propoxyphene hydrochloride and acetaminophen (91.6%), levothyroxine sodium (Levoxyl) (85.8%), atenolol (83.1%), and a combination of acetaminophen and codeine phosphate 3 (85.3%). The top 15 medications whose prices were most often underestimated or overestimated are shown in Figure 1, along with the percentage of correct responses for all physicians. No significant differences were observed between the sites.The mean (SD) score of the top 15 medications whose prices were most often underestimated or overestimated. Mean scores closer to zero indicate greater familiarity with the cost of the medication. Each unit on the x-axis represents one cost interval of $10. AQ indicates aqueous; MDI, metered dose inhaler; SR, sustained release; and XL, extended release.Table 1shows the mean scores and percentage of correct responses for all physicians and by faculty compared with residency class. For all physicians, the mean scores were significantly lower and the percentage of correct responses significantly higher for generic drugs and drugs categorized in the low-cost category, compared with branded drugs and drugs categorized in the medium- or high-cost category. This indicates greater familiarity by physicians with the cost of medications in these categories.Survey Respondents' Familiarity With Prescription Medication Costs*VariableType of MedicationAllBrandedGenericHigh CostMedium CostLow CostScore†All physicians1.51 (0.337)1.73 (0.474)1.20 (0.432)‡2.24 (0.882)1.40 (0.366)1.12 (0.477)§Faculty1.30 (0.288)&par;¶1.51 (0.397)&par;¶0.99 (0.361)&par;¶1.88 (0.777)¶1.29 (0.349)¶0.94 (0.369)&par;¶R3s1.43 (0.312)#1.60 (0.479)#1.21 (0.345)2.09 (0.886)#1.33 (0.391)1.11 (0.386)R2s1.58 (0.252)1.76 (1.375)**1.32 (0.454)2.25 (0.747)**1.47 (0.093)1.21 (0.540)R1s1.72 (0.344)2.02 (0.489)1.27 (0.486)2.75 (0.902)1.51 (0.398)1.21 (0.541)Correct responses, %All physicians22.9 (7.6)18.2 (8.3)30.6 (12.5)‡10.9 (12.6)21.9 (10.8)32.7 (14.0)§Faculty25.6 (7.4)¶19.9 (8.2)35.5 (13.1)¶15.0 (13.8)&par;¶22.6 (11.7)37.1 (14.7)¶R3s23.5 (9.0)20.2 (11.2)28.7 (11.4)13.2 (7.7)22.5 (11.9)31.7 (12.4)R2s21.9 (6.8)16.7 (6.5)30.1 (13.2)7.86 (9.2)20.8 (9.8)33.0 (14.7)R1s20.6 (6.1)15.9 (5.6)28.4 (11.3)7.69 (8.3)21.7 (10.0)29.2 (13.5)*Data are given as mean (SD). R3 indicates third-year resident; R2, second-year resident; and R1, first-year resident.†Scores closer to zero indicate greater familiarity with the cost of medication.‡tTest, P<.001 vs branded medications.§Analysis of variance, P<.05 for all comparisons between high-, medium-, and low-cost groups.&par;Analysis of variance, P<.05 for faculty vs R2s.¶Analysis of variance, P<.05 for faculty vs R1s.#Analysis of variance, P<.05 for R3s vs R1s.**Analysis of variance, P<.05 for R2s vs R1s.Faculty mean scores were significantly lower (indicating greater familiarity with costs of medications) than those of first- and second-year residents for all categories (P<.05 for all comparisons). Mean scores for third- and second-year residents were also significantly lower when compared with those of first-year residents for branded drugs (P≤.001 and P= .04, respectively). Faculty physicians were more likely to identify the correct price interval of all drugs (P= .008), generic drugs (P= .04), and drugs categorized into the low-cost (P= .04) or high-cost (P= .03) category compared with first-year residents. In all comparisons, no significant differences were observed between faculty and third-year residents. Linear regression revealed a trend for increasing years of practice correlating with lower mean scores for all drugs (r2= 0.11; P<.001) and for drugs categorized in the brand (r2= 0.05; P<.006), generic (r2= 0.08; P<.001), high-cost (r2= 0.03; P= .02), medium-cost (r2= 0.03; P= .02), and low-cost (r2= 0.04; P= .008) categories.Overall, 64.4% of the physicians believed that they were undereducated about the cost of the medications they prescribe, and 93.6% of the physicians believed that regular pricing information would help them prescribe more cost-effectively. Mean scores and the percentage of correct responses were not significantly (P>.05 for all comparisons) different between physicians reporting they received adequate cost information compared with those reporting they did not. Seventy-two percent of physicians reported that whether a patient has prescription drug insurance influences the choice of medication to prescribe; yet, they remembered to inquire about a patient's insurance status less than half (47.0%) of the time when writing a prescription.COMMENTDespite increased attention recently on increasing health care expenditures and prescription drug costs, family physicians are unfamiliar with the costs of common prescription drugs. More than two thirds of the responses (68.3%) underestimated the correct price interval. The average percentage of correct answers (22.9%) is consistent with reportspublished in the past 25 years.To our knowledge, this is the first study to address differences in prescription drug cost knowledge among faculty and resident physicians. First-year residents were least familiar with prescription medication costs, perhaps because of less experience in the drug therapy selection process. However, the survey was administered in March, which allowed first-year physicians nearly 8 months of experience in writing prescriptions. The small but statistically significant differences in mean scores between first- and third-year residents suggest that familiarity with prescription drug costs increases minimally during residency training. Once training is completed, this familiarity does not appear to increase, as evidenced by the lack of differences in scores between faculty and third-year residents and the small regression coefficient noted for years in practice and mean scores. If additional familiarity with medication costs is gained with practice, a stronger correlation between years graduated from medical school and correct responses would be expected. Most residency programs do not have formal education on prescription drug costs and methods to improve cost-effective prescribing. The small differences observed in pricing familiarity between faculty and residents suggest that formal education in this subject matter should be considered for inclusion in residency training curricula.Physicians in our study consistently underestimated the cost of branded drugs, while generic drugs were consistently overestimated. Glickman and colleaguesreported similar results in a survey of 132 primary care physicians on costs of medications commonly used in geriatric practice. Overestimation of brand drugs and underestimation of generic drugs suggest that physicians view medication costs within a narrow price range. The implication of the finding that generic drugs are consistently overestimated indicates physicians in many instances may not be aware that prescribing generic drugs in place of branded drugs (if available) can offer substantial cost savings to patients.Because prescription purchases are a directed demand to consumers, physicians should select generic alternatives if available and therapeutically indicated.Our survey is important because it assesses physician familiarity with medications they routinely prescribe. Previous studiesattempting to determine medication cost knowledge among physicians did not systematically identify drugs commonly prescribed by their surveyed populations. Lowy and colleaguesfirst documented lack of familiarity with drug costs among physicians. They cited insufficient attention to drug costs during training curricula and lack of cost information in professional articles and drug advertising as the primary reasons for physicians being unfamiliar with drug costs. Pricing information is not stressed in the pharmacology curricula of most medical schools. Other sources of cost information are limited.Rational, cost-effective prescribing depends on accurate and timely information about the cost of medications.In our study, most physicians reported that they did not receive sufficient information regarding cost of medications and nearly all believed regular updates on medication costs would help them prescribe more cost-effectively. Several studieshave demonstrated that education of physicians about drug prices can change prescribing behavior and reduce cost expenditures by improving selection of cost-effective therapies. Methods of educating physicians have included feedback reports on prescribing,educational programs on drug information and pricing,and academic detailing by clinical pharmacists.More recently, the effect of integrating on-line medication cost information in a computerized medical record system found no effect on reducing drug costs.These results suggest that educating physicians about prescription medication costs must be done actively within the context of providing information on therapeutic efficacy. Simply providing price lists is unlikely to alter prescribing habits.Several important limitations are noted in our study. The penetration of managed care into Iowa is limited; therefore, physicians in our survey may be less sensitive to medication costs than physicians in other states. Physicians in private practice, or those in specialties other than family medicine, may be more or less familiar with medication costs than the academic physicians in our study. Our findings, however, are consistent with surveys of other specialties and of those in private practice. This suggests that the lack of pricing knowledge is homogeneous across disciplines and practice sites. Our prescription audit occurred in January, the peak month for respiratory tract infections. This might skew our description of the most commonly prescribed drugs in favor of antibiotics. However, we audited more than 3500 prescriptions in an attempt to reduce this bias, and only 20% of the medications on the survey were antibiotics. In addition, we believe the medication list distributed in the survey accurately represents common medications prescribed nationally. The medications included on the survey are consistent with data from the National Center for Health Statistics 1996 National Ambulatory Medical Care Survey on most common drug mentions by therapeutic classification at physician office visits.Last, not all of the medications in our survey were priced at the midpoint of each pricing interval. This could lead to potential misclassifications when medication costs lie at the extremes of the interval. As the actual drug price increases, the magnitude of this effect is diminished. The broad range of medication costs (<$10->$100) in our survey reduces this potential bias.Despite limitations to survey research, our study has several strengths, including the high response rate (86.8%), addressing differences in prescription drug cost familiarity between faculty and resident physicians, and assessing familiarity with medications routinely prescribed. Our scoring system of mean scores was unique and allowed us to examine the magnitude of the variation in responses. Likewise, we classified medications into $10 price intervals, rather than asking for specific cost; price ranges are more representative of how physicians think of medication costs when making prescribing decisions. Although we surveyed physicians from residency programs, we believe our sample is representative of family physicians elsewhere, because our findings are consistent with those that have included family physicians in private practice.In summary, physicians in our study were unfamiliar with the costs of medications they commonly prescribe, and reported they do not receive adequate information on costs of medications. Nearly all believe receiving regular pricing information would help them prescribe more cost-effectively. To help control escalating drug costs and assist physicians in selecting cost-effective therapies, future studies should evaluate the role and impact on prescribing of didactic educational programs (which address the costs of prescription medications) incorporated into residency training curricula.SSmithMFreelandSHefflerDMcKusickthe Health Expenditures Projection TeamThe next ten years of health spending: what does the future hold?Health Aff (Millwood).1998;17:128-140.SMSchappertAmbulatory care visits to physician offices, hospital outpatient departments, and emergency departments: United States, 1996.Vital Health Stat 13.1998:1-37.DHKrelingDAMottJBWiederholtJLundyLLevittPrescription Drug Trends: A Chartbook: Final Report.Menlo Park, Calif: Kaiser Family Foundation; 2000. Report 3019.LGlickmanEABruceFGCaroJAvornPhysicians' knowledge of drug costs for the elderly.J Am Geriatr Soc.1994;42:992-996.NBShulmanBMartinezDBroganFinancial cost as an obstacle to hypertension therapy.Am J Public Health.1986;76:1105-1108.JEHuxCDNaylorDrug prices and third party payment: do they influence medication selection?Pharmacoeconomics.1994;5:343-350.MMReidenbergA plea for prices in physician prescribing [letter].JAMA.1991;266:3285.DRLowyLLowyRSWarnerA survey of physicians' awareness of drug costs.J Med Educ.1972;47:349-351.JLFinkDJKerriganPhysicians' knowledge of drug prices.Contemp Pharm Pract.1978;1:18-21.MLWeberCAugerRClerouxKnowledge of medical students, pediatric residents, and pediatricians about the cost of some medications.Pediatr Pharmacol (New York).1986;5:281-285.LGMillerABlumPhysician awareness of prescription drug costs: a missing element of drug advertising and promotion.J Fam Pract.1993;36:33-36.DWalzakSSwindellsABhardwajPrimary care physicians and the cost of drugs: a study of prescribing practices based on recognition and information sources.J Clin Pharmacol.1994;34:1159-1163.JHoffmanFABarefieldSRamamurthyA survey of physician knowledge of drug costs.J Pain Symptom Manage.1995;10:432-435.MWGriswoldLLBricelandTSLesarCosts of therapeutic alternatives: physicians' knowledge [letter].Ann Pharmacother.1996;30:882-883.DAWoodwellNational Ambulatory Medical Care Survey: 1997 Summary.Hyattsville, Md: National Center for Health Statistics; 1999. Advance Data From Vital and Health Statistics, No. 305.HMGarrettRed Book.Montvale, NJ: Medical Economics Books; 1998.ADBowerGLBurkettFamily physician and generic drugs: a study of recognition, information sources, prescribing attitudes, and practices.J Fam Pract.1987;24:612-616.LMFrazierJTBrownGWDivineCan physician education lower the cost of prescription drugs?Ann Intern Med.1991;115:116-121.JESumptonTCFrewenMJRiederThe effect of physician education on knowledge of drug therapeutics and costs.Ann Pharmacother.1992;26:692-697.JAvornSBSoumeraiImproving drug-therapy decisions through educational outreach: a randomized controlled trial of academically based "detailing."N Engl J Med.1983;308:1457-1463.SMOrnsteinLLMacFarlaneRGJenkinsMedication cost information in a computer-based patient record system.Arch Fam Med.1999;8:118-121.MRyanBYuleCBondRJTaylorScottish general practitioners' attitudes and knowledge in respect of prescribing costs.BMJ.1990;300:1316-1318.MRyanBYuleCBondRTaylorKnowledge of drug costs: a comparison of general practitioners in Scotland and England.Br J Gen Pract.1992;42:6-9.Accepted for publication August 30, 2000.Presented in part as a poster at the American College of Clinical Pharmacy Annual Meeting, Cincinnati, Ohio, November 9, 1998.We thank Erika J. Ernst, PharmD, for statistical consultation during the preparation of the manuscript.Drs Ernst, Kelly, Hoehns, Swegle, Buys, Logemann, Ford, Kautzman, and Pretorius participated in the original idea and design of the survey; Drs Ernst, Kelly, Hoehns, Swegle, Buys, Logemann, Ford, and Kautzman participated in the data collection; Drs Ernst, Kelly, and Sorofman participated in the data analysis; and all of the authors participated in the write-up and review of the manuscript.Corresponding author: Michael E. Ernst, PharmD, College of Pharmacy, The University of Iowa, S411 Pharmacy Bldg, Iowa City, IA 52242 (e-mail: michael-ernst@uiowa.edu). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Archives of Family Medicine American Medical Association

Loading next page...
 
/lp/american-medical-association/prescription-medication-costs-9Ga9OWqe2R

References (4)

Publisher
American Medical Association
Copyright
Copyright 2000 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
ISSN
1063-3987
eISSN
1063-3987
DOI
10-1001/pubs.Arch Fam Med.-ISSN-1063-3987-9-10-foc0029
Publisher site

Abstract

BackgroundStudies in the past 25 years have suggested that physicians are not familiar with the costs of common prescription medications.ObjectivesTo determine physician familiarity with the cost of common prescription medications and to determine the value physicians place on knowing information regarding the cost of medications.DesignSurvey.SettingSeven community-based family medicine residency teaching clinics in Iowa.ParticipantsTwo hundred five practicing resident and faculty physicians.InterventionsFrom a series of $10 price intervals (range, $0.01-$80.00), physicians were asked to select the interval containing the cash price of the medication to an uninsured patient for 50 medications commonly prescribed in outpatient family medicine clinics. Physicians were also questioned about the value of medication cost information to their practice.Main Outcome MeasuresThe percentage of correct responses and the mean pricing scores were calculated for each respondent and for all medications.ResultsOne hundred seventy-eight physicians responded (86.8%). Only 22.9% of the responses correctly identified the cost of the medication. More than two thirds (68.3%) of the responses underestimated the correct price interval. Branded drugs were underestimated in 89.9% of responses, while generic drugs were overestimated in 90.2% of responses. Overall, 64.4% of physicians believed they did not receive sufficient information in their practices regarding prescription drug costs, and nearly all (93.6%) reported that regular information on prescription medication costs would help them prescribe more cost-effectively.ConclusionsPhysicians are unfamiliar with the costs of medications they commonly prescribe, and they report that regular access to information on prescription medication costs would help them prescribe more cost-effectively.INCREASED SPENDING and use of prescription medications contribute to escalating costs of health care. Attempts to control drug costs, through implementation of formularies and mandatory generic substitution, have had little impact on reducing the economic burden of prescription drugs. In the next 10 years, spending on prescription drugs is anticipated to increase an average of 9.8%, and account for nearly 8% of all national health expenditures.With 60% of physician office visits resulting in a prescription being written,selection of cost-effective pharmacotherapy is critical to curb increasing drug costs.On average, Americans use about 10 prescriptions per year.Access to affordable medications remains limited for some patients. Medicare does not have a prescription drug benefit, and approximately 30% of prescriptions filled are purchased by uninsured patients.In one survey,85% of physicians reported that inability to afford medications was problematic for their patients. Drug costs clearly influence patient compliance with therapy, and the inability to pay for medications can adversely affect treatment outcomes.While prescription drug insurance may reduce costs for individual patients, the inappropriate selection of more expensive medications may actually increase overall costs to society.Patients with prescription drug insurance are known to use brand name drugs at a higher rate than those who pay cash, which may reflect, in part, the fact that prescription coverage helps to insulate these consumers from the additional costs of more expensive brand name drugs.The ability of the physician to select cost-effective therapy is, therefore, important not only for the patient but also for consumers and employers paying the cost of the health insurance as well.Prescription drug purchases are economically unique because they are a "directed" demand.Physicians direct the purchase through drug selection and determination of appropriateness, with minimal input from the consumer. As the primary target of prescription drug marketing, physicians influence the market by acting as "gatekeepers" who direct prescription drug use decisions. To select cost-effective therapies, physicians must have an adequate knowledge of prescription drug costs.Studiesover the past 25 years have documented that physicians are not familiar with prescription drug costs. These studies have limited generalizability, as the method by which the drugs were selected does not ensure physicians surveyed routinely prescribe those medications. The most common office visit is to a family physician, and family physicians write more than 30% of prescriptions annually in the United States.Family physicians are thus in a position to direct most prescription drug purchases, and should have a basic familiarity with the costs of medications they routinely prescribe. This study identifies commonly prescribed medications in family physician clinics and determines the familiarity of physicians with the costs of these medications and their views on the importance of medication cost information to their prescribing behavior.PARTICIPANTS AND METHODSSTUDY PARTICIPANTSPracticing faculty and resident physicians from 7 family medicine residency training programs geographically distributed throughout Iowa participated. All programs are community based, in towns ranging from 30 000 to 190 000 people, averaging 15 000 to 40 000 patient visits per year; there is wide diversity of patients from each site. The approximate insurance status is as follows: Medicaid, 15% (range, 10%-19%); Medicare, 15% (range, 7%-25%); managed care, 23% (range, 10%-30%); and uninsured, 25% (range, 6%-56%). Each residency program has an average of 20 to 30 residents, 8 to 12 faculty physicians, and 1 clinical pharmacist. Residents at each site spend time in the ambulatory setting according to guidelines set by the American Council of Graduate Medical Education Residency Review Committee for Family Practice. First-year residents spend at least half a day per week in the ambulatory outpatient clinic, while second- and third-year residents spend at least 2 and 3 half days per week in the outpatient clinic, respectively.STUDY PROCEDURESDuring January 1998, a daily audit of written prescriptions was performed by clinical pharmacist faculty members (M.E.E., M.W.K., J.D.H., J.M.S., L.M.B., C.D.L., J.K.F., and H.A.K.) at each site. A representative sample of 50 commonly prescribed drugs among all clinics was selected for the survey, from 423 different medications and 3635 total prescriptions audited. For each medication on the survey, an attempt was made to insert into the survey at least one comparator medication from the same therapeutic class that was also in the top 50 commonly prescribed agents from the audit. The final 50 drugs chosen for the survey represented 50.3% of all prescriptions audited during the study period.Medications were presented in random order on the survey to prevent any cost patterns from emerging, and quantities were chosen to approximate a 1-month supply for long-term medications or the standard course of therapy for short-term medications. Generic names were used, with brand names supplied for medications available only as single source, branded drugs. Physicians were instructed to assume the prescription would be dispensed as written, and then asked to select what they believed to be the correct prescription cost range for an uninsured patient from a series of $10 price intervals (range, $0.01-$80). Physicians were also asked to respond on a Likert scale to 4 statements regarding the importance of medication cost information to their practice. Statements included asking whether physicians believed they received adequate information regarding the costs of the medications they prescribe, whether regularly accessible information on medication costs would help improve their practice and help them prescribe more cost-effectively, and whether the prescription drug insurance status of patients influences their choice of medication to prescribe. The survey was distributed to faculty and residents during March 1998, which allowed first-year residents to gain almost 8 months of prescribing experience in the outpatient setting. The survey took approximately 15 minutes to complete.Because medication prices can vary greatly by type of pharmacy (chain or independent) and location, both of which could affect the scoring of the survey, the actual price for each medication was calculated using a standard formula used by Iowa Medicaid for reimbursement: branded drugs = [average wholesale price (published in the 1998 Red Book) − 10%] + $6.25, and generic drugs = maximum allowable cost set by the Health Care Financing Administration for multisource drugs + $4.02. Two local pharmacies were surveyed to compare the calculated price vs the price charged by the pharmacy. Discrepancies of more than $10 were resolved by using the pharmacy-reported price. This occurred in less than 6 cases. Correct price interval selections were coded with a value of zero. Incorrect selections were coded as the absolute value of the number of intervals above or below the correct interval.STATISTICAL ANALYSISAll statistical analyses were performed using statistical software (Statistical Product and Service Solutions, version 7.5.1; SPSS Inc, Chicago, Ill). Drugs were categorized as branded or generic, and as high (≥$60.01), medium ($20.01-$60.00), and low ($0.01-$20.00) cost for comparisons.The total number and the percentage of correct responses, as well as mean scores, were calculated for each respondent and for all medications. Mean scores were calculated to allow for more accurate reporting of the magnitude of the difference between price estimated by the physician and correct price interval. Because absolute values were used to code the number of price intervals above or below the correct interval, mean scores closer to zero indicate greater familiarity with price. Paired ttests were performed to identify differences in mean scores, and in percentage of correct responses for branded and generic categories. An analysis of variance with post-hoc Tukey correction for multiple comparisons was used to identify differences in mean scores and percentage of correct responses for each category, and for comparisons between faculty and resident classes. P<.05 was considered statistically significant.RESULTSThe survey was completed by 178 (86.8%) of 205 eligible physicians. Nonresponders to the survey generally were physicians who were on an off-site rotation at the time the survey was administered. Respondents were nearly equally distributed between faculty (n = 43) and first-year (n = 44), second-year (n = 46), and third-year (n = 45) residents. For faculty respondents, the mean (±SD) number of years since graduation from medical school was 17.4 (±7.9) years, with a mean (±SD) of 8.6 (±6.5) years as faculty.Overall, 68.3% of the responses underestimated the correct price interval for all drugs. Only 22.9% of the responses correctly identified the cost. The prices of branded drugs were underestimated in 89.9% of responses, while generic drugs were overestimated in 90.2% of responses. The medications whose prices were most commonly underestimated included fluoxetine hydrochloride (Prozac) (93.8% of responses underestimated), cefuroxime axetil (Ceftin) (82.6%), ciprofloxacin hydrochloride (Cipro) (87.0%), loratadine (Claritin) (87.0%), and omeprazole (Prilosec) (76.8%), while the medications most commonly overestimated included trazodone hydrochloride (94.4% of responses overestimated), a combination of propoxyphene hydrochloride and acetaminophen (91.6%), levothyroxine sodium (Levoxyl) (85.8%), atenolol (83.1%), and a combination of acetaminophen and codeine phosphate 3 (85.3%). The top 15 medications whose prices were most often underestimated or overestimated are shown in Figure 1, along with the percentage of correct responses for all physicians. No significant differences were observed between the sites.The mean (SD) score of the top 15 medications whose prices were most often underestimated or overestimated. Mean scores closer to zero indicate greater familiarity with the cost of the medication. Each unit on the x-axis represents one cost interval of $10. AQ indicates aqueous; MDI, metered dose inhaler; SR, sustained release; and XL, extended release.Table 1shows the mean scores and percentage of correct responses for all physicians and by faculty compared with residency class. For all physicians, the mean scores were significantly lower and the percentage of correct responses significantly higher for generic drugs and drugs categorized in the low-cost category, compared with branded drugs and drugs categorized in the medium- or high-cost category. This indicates greater familiarity by physicians with the cost of medications in these categories.Survey Respondents' Familiarity With Prescription Medication Costs*VariableType of MedicationAllBrandedGenericHigh CostMedium CostLow CostScore†All physicians1.51 (0.337)1.73 (0.474)1.20 (0.432)‡2.24 (0.882)1.40 (0.366)1.12 (0.477)§Faculty1.30 (0.288)&par;¶1.51 (0.397)&par;¶0.99 (0.361)&par;¶1.88 (0.777)¶1.29 (0.349)¶0.94 (0.369)&par;¶R3s1.43 (0.312)#1.60 (0.479)#1.21 (0.345)2.09 (0.886)#1.33 (0.391)1.11 (0.386)R2s1.58 (0.252)1.76 (1.375)**1.32 (0.454)2.25 (0.747)**1.47 (0.093)1.21 (0.540)R1s1.72 (0.344)2.02 (0.489)1.27 (0.486)2.75 (0.902)1.51 (0.398)1.21 (0.541)Correct responses, %All physicians22.9 (7.6)18.2 (8.3)30.6 (12.5)‡10.9 (12.6)21.9 (10.8)32.7 (14.0)§Faculty25.6 (7.4)¶19.9 (8.2)35.5 (13.1)¶15.0 (13.8)&par;¶22.6 (11.7)37.1 (14.7)¶R3s23.5 (9.0)20.2 (11.2)28.7 (11.4)13.2 (7.7)22.5 (11.9)31.7 (12.4)R2s21.9 (6.8)16.7 (6.5)30.1 (13.2)7.86 (9.2)20.8 (9.8)33.0 (14.7)R1s20.6 (6.1)15.9 (5.6)28.4 (11.3)7.69 (8.3)21.7 (10.0)29.2 (13.5)*Data are given as mean (SD). R3 indicates third-year resident; R2, second-year resident; and R1, first-year resident.†Scores closer to zero indicate greater familiarity with the cost of medication.‡tTest, P<.001 vs branded medications.§Analysis of variance, P<.05 for all comparisons between high-, medium-, and low-cost groups.&par;Analysis of variance, P<.05 for faculty vs R2s.¶Analysis of variance, P<.05 for faculty vs R1s.#Analysis of variance, P<.05 for R3s vs R1s.**Analysis of variance, P<.05 for R2s vs R1s.Faculty mean scores were significantly lower (indicating greater familiarity with costs of medications) than those of first- and second-year residents for all categories (P<.05 for all comparisons). Mean scores for third- and second-year residents were also significantly lower when compared with those of first-year residents for branded drugs (P≤.001 and P= .04, respectively). Faculty physicians were more likely to identify the correct price interval of all drugs (P= .008), generic drugs (P= .04), and drugs categorized into the low-cost (P= .04) or high-cost (P= .03) category compared with first-year residents. In all comparisons, no significant differences were observed between faculty and third-year residents. Linear regression revealed a trend for increasing years of practice correlating with lower mean scores for all drugs (r2= 0.11; P<.001) and for drugs categorized in the brand (r2= 0.05; P<.006), generic (r2= 0.08; P<.001), high-cost (r2= 0.03; P= .02), medium-cost (r2= 0.03; P= .02), and low-cost (r2= 0.04; P= .008) categories.Overall, 64.4% of the physicians believed that they were undereducated about the cost of the medications they prescribe, and 93.6% of the physicians believed that regular pricing information would help them prescribe more cost-effectively. Mean scores and the percentage of correct responses were not significantly (P>.05 for all comparisons) different between physicians reporting they received adequate cost information compared with those reporting they did not. Seventy-two percent of physicians reported that whether a patient has prescription drug insurance influences the choice of medication to prescribe; yet, they remembered to inquire about a patient's insurance status less than half (47.0%) of the time when writing a prescription.COMMENTDespite increased attention recently on increasing health care expenditures and prescription drug costs, family physicians are unfamiliar with the costs of common prescription drugs. More than two thirds of the responses (68.3%) underestimated the correct price interval. The average percentage of correct answers (22.9%) is consistent with reportspublished in the past 25 years.To our knowledge, this is the first study to address differences in prescription drug cost knowledge among faculty and resident physicians. First-year residents were least familiar with prescription medication costs, perhaps because of less experience in the drug therapy selection process. However, the survey was administered in March, which allowed first-year physicians nearly 8 months of experience in writing prescriptions. The small but statistically significant differences in mean scores between first- and third-year residents suggest that familiarity with prescription drug costs increases minimally during residency training. Once training is completed, this familiarity does not appear to increase, as evidenced by the lack of differences in scores between faculty and third-year residents and the small regression coefficient noted for years in practice and mean scores. If additional familiarity with medication costs is gained with practice, a stronger correlation between years graduated from medical school and correct responses would be expected. Most residency programs do not have formal education on prescription drug costs and methods to improve cost-effective prescribing. The small differences observed in pricing familiarity between faculty and residents suggest that formal education in this subject matter should be considered for inclusion in residency training curricula.Physicians in our study consistently underestimated the cost of branded drugs, while generic drugs were consistently overestimated. Glickman and colleaguesreported similar results in a survey of 132 primary care physicians on costs of medications commonly used in geriatric practice. Overestimation of brand drugs and underestimation of generic drugs suggest that physicians view medication costs within a narrow price range. The implication of the finding that generic drugs are consistently overestimated indicates physicians in many instances may not be aware that prescribing generic drugs in place of branded drugs (if available) can offer substantial cost savings to patients.Because prescription purchases are a directed demand to consumers, physicians should select generic alternatives if available and therapeutically indicated.Our survey is important because it assesses physician familiarity with medications they routinely prescribe. Previous studiesattempting to determine medication cost knowledge among physicians did not systematically identify drugs commonly prescribed by their surveyed populations. Lowy and colleaguesfirst documented lack of familiarity with drug costs among physicians. They cited insufficient attention to drug costs during training curricula and lack of cost information in professional articles and drug advertising as the primary reasons for physicians being unfamiliar with drug costs. Pricing information is not stressed in the pharmacology curricula of most medical schools. Other sources of cost information are limited.Rational, cost-effective prescribing depends on accurate and timely information about the cost of medications.In our study, most physicians reported that they did not receive sufficient information regarding cost of medications and nearly all believed regular updates on medication costs would help them prescribe more cost-effectively. Several studieshave demonstrated that education of physicians about drug prices can change prescribing behavior and reduce cost expenditures by improving selection of cost-effective therapies. Methods of educating physicians have included feedback reports on prescribing,educational programs on drug information and pricing,and academic detailing by clinical pharmacists.More recently, the effect of integrating on-line medication cost information in a computerized medical record system found no effect on reducing drug costs.These results suggest that educating physicians about prescription medication costs must be done actively within the context of providing information on therapeutic efficacy. Simply providing price lists is unlikely to alter prescribing habits.Several important limitations are noted in our study. The penetration of managed care into Iowa is limited; therefore, physicians in our survey may be less sensitive to medication costs than physicians in other states. Physicians in private practice, or those in specialties other than family medicine, may be more or less familiar with medication costs than the academic physicians in our study. Our findings, however, are consistent with surveys of other specialties and of those in private practice. This suggests that the lack of pricing knowledge is homogeneous across disciplines and practice sites. Our prescription audit occurred in January, the peak month for respiratory tract infections. This might skew our description of the most commonly prescribed drugs in favor of antibiotics. However, we audited more than 3500 prescriptions in an attempt to reduce this bias, and only 20% of the medications on the survey were antibiotics. In addition, we believe the medication list distributed in the survey accurately represents common medications prescribed nationally. The medications included on the survey are consistent with data from the National Center for Health Statistics 1996 National Ambulatory Medical Care Survey on most common drug mentions by therapeutic classification at physician office visits.Last, not all of the medications in our survey were priced at the midpoint of each pricing interval. This could lead to potential misclassifications when medication costs lie at the extremes of the interval. As the actual drug price increases, the magnitude of this effect is diminished. The broad range of medication costs (<$10->$100) in our survey reduces this potential bias.Despite limitations to survey research, our study has several strengths, including the high response rate (86.8%), addressing differences in prescription drug cost familiarity between faculty and resident physicians, and assessing familiarity with medications routinely prescribed. Our scoring system of mean scores was unique and allowed us to examine the magnitude of the variation in responses. Likewise, we classified medications into $10 price intervals, rather than asking for specific cost; price ranges are more representative of how physicians think of medication costs when making prescribing decisions. Although we surveyed physicians from residency programs, we believe our sample is representative of family physicians elsewhere, because our findings are consistent with those that have included family physicians in private practice.In summary, physicians in our study were unfamiliar with the costs of medications they commonly prescribe, and reported they do not receive adequate information on costs of medications. Nearly all believe receiving regular pricing information would help them prescribe more cost-effectively. To help control escalating drug costs and assist physicians in selecting cost-effective therapies, future studies should evaluate the role and impact on prescribing of didactic educational programs (which address the costs of prescription medications) incorporated into residency training curricula.SSmithMFreelandSHefflerDMcKusickthe Health Expenditures Projection TeamThe next ten years of health spending: what does the future hold?Health Aff (Millwood).1998;17:128-140.SMSchappertAmbulatory care visits to physician offices, hospital outpatient departments, and emergency departments: United States, 1996.Vital Health Stat 13.1998:1-37.DHKrelingDAMottJBWiederholtJLundyLLevittPrescription Drug Trends: A Chartbook: Final Report.Menlo Park, Calif: Kaiser Family Foundation; 2000. Report 3019.LGlickmanEABruceFGCaroJAvornPhysicians' knowledge of drug costs for the elderly.J Am Geriatr Soc.1994;42:992-996.NBShulmanBMartinezDBroganFinancial cost as an obstacle to hypertension therapy.Am J Public Health.1986;76:1105-1108.JEHuxCDNaylorDrug prices and third party payment: do they influence medication selection?Pharmacoeconomics.1994;5:343-350.MMReidenbergA plea for prices in physician prescribing [letter].JAMA.1991;266:3285.DRLowyLLowyRSWarnerA survey of physicians' awareness of drug costs.J Med Educ.1972;47:349-351.JLFinkDJKerriganPhysicians' knowledge of drug prices.Contemp Pharm Pract.1978;1:18-21.MLWeberCAugerRClerouxKnowledge of medical students, pediatric residents, and pediatricians about the cost of some medications.Pediatr Pharmacol (New York).1986;5:281-285.LGMillerABlumPhysician awareness of prescription drug costs: a missing element of drug advertising and promotion.J Fam Pract.1993;36:33-36.DWalzakSSwindellsABhardwajPrimary care physicians and the cost of drugs: a study of prescribing practices based on recognition and information sources.J Clin Pharmacol.1994;34:1159-1163.JHoffmanFABarefieldSRamamurthyA survey of physician knowledge of drug costs.J Pain Symptom Manage.1995;10:432-435.MWGriswoldLLBricelandTSLesarCosts of therapeutic alternatives: physicians' knowledge [letter].Ann Pharmacother.1996;30:882-883.DAWoodwellNational Ambulatory Medical Care Survey: 1997 Summary.Hyattsville, Md: National Center for Health Statistics; 1999. Advance Data From Vital and Health Statistics, No. 305.HMGarrettRed Book.Montvale, NJ: Medical Economics Books; 1998.ADBowerGLBurkettFamily physician and generic drugs: a study of recognition, information sources, prescribing attitudes, and practices.J Fam Pract.1987;24:612-616.LMFrazierJTBrownGWDivineCan physician education lower the cost of prescription drugs?Ann Intern Med.1991;115:116-121.JESumptonTCFrewenMJRiederThe effect of physician education on knowledge of drug therapeutics and costs.Ann Pharmacother.1992;26:692-697.JAvornSBSoumeraiImproving drug-therapy decisions through educational outreach: a randomized controlled trial of academically based "detailing."N Engl J Med.1983;308:1457-1463.SMOrnsteinLLMacFarlaneRGJenkinsMedication cost information in a computer-based patient record system.Arch Fam Med.1999;8:118-121.MRyanBYuleCBondRJTaylorScottish general practitioners' attitudes and knowledge in respect of prescribing costs.BMJ.1990;300:1316-1318.MRyanBYuleCBondRTaylorKnowledge of drug costs: a comparison of general practitioners in Scotland and England.Br J Gen Pract.1992;42:6-9.Accepted for publication August 30, 2000.Presented in part as a poster at the American College of Clinical Pharmacy Annual Meeting, Cincinnati, Ohio, November 9, 1998.We thank Erika J. Ernst, PharmD, for statistical consultation during the preparation of the manuscript.Drs Ernst, Kelly, Hoehns, Swegle, Buys, Logemann, Ford, Kautzman, and Pretorius participated in the original idea and design of the survey; Drs Ernst, Kelly, Hoehns, Swegle, Buys, Logemann, Ford, and Kautzman participated in the data collection; Drs Ernst, Kelly, and Sorofman participated in the data analysis; and all of the authors participated in the write-up and review of the manuscript.Corresponding author: Michael E. Ernst, PharmD, College of Pharmacy, The University of Iowa, S411 Pharmacy Bldg, Iowa City, IA 52242 (e-mail: michael-ernst@uiowa.edu).

Journal

Archives of Family MedicineAmerican Medical Association

Published: Nov 1, 2000

There are no references for this article.